PSN-L Email List Message
Subject: Re: Re: Basic Programming Help desired?
From: "Geoff" gmvoeth@...........
Date: Fri, 9 May 2008 14:28:49 -0700
Hello Cruel World;
I have learned by simply changing my arrays from
Virtual ( EMS MEMORY ) to HUGE ( REAL MODE UNDER 1MB )
all my routines relating to arrays is faster by like
17X faster. There is something rotten in denmark
in regards to memory management over 1 MB.
I got this fft routine that now completes
128 lines of results in one second instead of 17 seconds.
Now I have this PowerBasic for Windows that
I have never tried but understand is faster code
but it will not run under DOS as i remember and
you need to make DLLs and the like which
are foreign to my way of thinking.
Why the heck dont the world of science
beat the world of business into a corner and
create a decent operating system for
science oriented persons.
We do not need all that security garbage like the
money people do ( meaning Protected MODE ).
----- Original Message -----
From: "Kevin Brunt"
Sent: Thursday, May 08, 2008 2:50 PM
Subject: Fwd: Re: Basic Programming Help desired?
> I think that you'll find that PUTting a 32768 byte array will be broadly comparable in time to BSAVEing the same array.
> The timing issues that you have been having are almost certainly due to the manipulation of individual array elements. Your first
> program was appending each byte to the end of a string one-by-one; I suspect that appending a byte to an N byte string consists of
> creating a new N+1 byte string, copying the N byte string to it and adding the new byte to the end - the original N byte string is
> then discarded. The observed delays are due to the repeated copying.
> Your later program wrote one byte at a time to the file; since there will be a (largely) fixed overhead for each call to PUT, 2^15
> calls to PUT takes a lot longer than one call.
> Incidentally, I've been lurking in the PSN-L archives because I've been working on a PDP-11 (in a museum) that used to be used for
> seismic processing. The equivalent program to yours in PDP-11 assembler (for the RT-11 operating system) is probably no longer
> than your BASIC programs, and RT-11 can perform the write-to-disk while the program is filling up the next buffer. (Admittedly,
> the PDP-11 has 4 6-foot high 19-inch cabinets, each with its own mains lead....)
>>Date: Thu, 8 May 2008 13:55:50 -0700
>>Hello Mr. Brunt ;
>>Yes you are basically correct.
>>But speed of saving the array is the real issue.
>>I have tried just about everything but found
>>nothing faster than using the BSAVE command.
>>ps; I never tried to put a complete array in a single
>>operation. Did not know that was possible.
>>I still think BSAVE will be faster.
> Public Seismic Network Mailing List (PSN-L)
> To leave this list email PSN-L-REQUEST@.............. with
> the body of the message (first line only): unsubscribe
> See http://www.seismicnet.com/maillist.html for more information.
Public Seismic Network Mailing List (PSN-L)
[ Top ]
[ Back ]
[ Home Page ]